- Top of page
Summary: Purpose: Whether status epilepticus (SE) in early infancy, rather than the underlying illness, leads to temporal lobe neurodegeneration and volume reduction remains controversial.
Methods: SE was induced with LiCl-pilocarpine in P12 rats. To assess acute neuronal damage, brains (five controls, five with SE) were investigated at 8 h after SE by using silver and Fluoro-Jade B staining. Some brains from the early phase were processed for electron microscopy. To assess chronic changes, brains from nine controls and 13 rats with SE at P12 were analyzed after 3 months by using histology and magnetic resonance imaging (MRI).
Results: MRI analysis of the temporal lobe of adult animals with SE at P12 indicated that 23% of the rats had hippocampal, 15% had amygdaloid, and 31% had perirhinal volume reduction. Histologic analysis of sections from the MR-imaged brains correlated with the MRI data. Analysis of neurodegeneration 8 h after SE by using both silver and Fluoro-Jade B staining revealed degenerating neurons located in the same temporal lobe regions as the volume reduction in chronic samples. Electron microscopic analysis revealed irreversible ultrastructural alterations. As with the chronic histologic and MRI findings, interanimal variability was seen in the distribution and severity of acute damage.
Conclusions: Our data indicate that SE at P12 can cause acute neurodegeneration in the hippocampus as well as in the adjacent temporal lobe. It is likely that acute neuronal death contributes to volume reduction in temporal lobe regions that is detected with MRI in a subpopulation of animals in adulthood.
Status epilepticus (SE) is a neurologic emergency with a higher incidence in infancy and childhood than in any other period of life (1). It remains controversial, however, whether SE causes injury to the developing brain. Prospective imaging studies demonstrated volume reduction of the hippocampus between the two consecutive measurements in a subpopulation of infants and children with prolonged febrile seizures (2). Other studies, however, suggest that the association between prolonged seizures and structural abnormalities can result from complex interactions between developmental abnormalities, prenatal or perinatal insults, and genetic factors [for review, see (3)].
In clinical studies, the causality between seizure activity and structural damage is difficult to investigate because of associated illnesses and treatments. Experimental studies in which SE was induced in normal immature brain demonstrated that SE can cause hippocampal and amygdaloid neurodegeneration resembling that in human temporal lobe epilepsy (TLE) in rats older than postnatal day (P) 14 (4). We recently extended these observations by showing that convulsive SE for 2 h can cause neuronal death in the mediodorsal nucleus of the thalamus in 100% of animals as early as P12 (5). Further, rats with SE at P12 had impaired memory and emotional behavior when assessed 3 months later, and ∼25% of animals developed spontaneous seizures within 3 to 6 months, which involved the hippocampus (6). These findings suggest that SE as early as P12, corresponding to early infancy in humans, is associated with structural alterations with an unfavorable functional outcome in a subpopulation of animals. It is not known, however, whether SE at P12 causes the volume reduction of the medial temporal lobe that is a typical feature of pathology in TLE patients with early-life SE.
We designed a study to address four questions. First, can volume decrease be detected in hippocampus, amygdala, or the surrounding cortex in adult rats with SE at P12? Second, is neurodegeneration present in all animals or only in a subpopulation? Third, does the volume decrease revealed by magnetic resonance imaging (MRI) correlate with neurodegeneration assessed from histologic sections from the same animals? Fourth, does the chronic volume reduction locate in the same regions as acute neurodegeneration?
SE was induced with LiCl-pilocarpine in P12 rats. In one group, neurodegeneration was studied 8 h after SE by using Fluoro-Jade B and silver staining. A second group of animals was allowed to survive for 3 months, after which they underwent MRI and histologic analysis. Our data indicate that SE at P12 can cause the acute neurodegeneration in the hippocampus, as well as in the adjacent temporal lobe, detected with MRI in a subpopulation of animals in adulthood.
- Top of page
We investigated whether experimentally induced SE in rat during early postnatal life leads to volume reduction of the temporal lobe, which is a typical finding in epilepsy patients with SE in childhood. Neurodegeneration was assessed by using histology and MRI. Three major findings occurred in the present study. First, MRI revealed volume decline in the hippocampus, amygdala, and the perirhinal cortex in adult rats with SE at P12. Second, volume decrease was apparent only in a subpopulation of animals. Third, volume reduction detected in MRI 3 months after SE appeared in the same regions as the degenerating neurons in preparations 8 h after SE. Irreversibility of early neuronal injury was confirmed by electron microscopy.
Experimental models of SE provide several advantages over clinical studies to examine the structural and functional consequences of SE in developing brain. For example, the study population can be made more homogeneous regarding genetic background, age, duration of SE, follow-up, and outcome measures, yet the severity and distribution of degenerating cells remains variable between animals, as described in the present histologic and MRI analysis, and which has also been shown in other studies (13,14). Such details are often obscure because the results are presented as mean values of a group of animals, and interanimal variability in pathology is rarely discussed, although it might be a critical factor explaining the variability in functional outcome after SE (i.e., duration of the epileptogenic period or seizure frequency and seizure type in animals that develop epilepsy after SE) (6,15).
Duration of SE is one of the major factors determining the severity of neurodegeneration (16). To standardize the duration of SE and reduce variability, we administered paraldehyde at 2 h to stop SE. Our recent video-EEG study conducted in parallel with the present analysis demonstrated that paraldehyde administration efficiently suppresses behavioral seizures (6). Electrographic seizures were, however, only transiently (<2 h) suppressed. Short ictal episodes in EEG could be recorded for ≤10 to 16 h after the beginning of SE (6). Therefore we cannot exclude the possibility that interindividual variability both in the acute and chronic animal groups relates to variability in the duration and severity of SE, despite paraldehyde administration. These data strengthen the view that despite technical difficulties, the quantification of SE with EEG in P12 rats provides valuable information about the association of duration and severity of SE with neurodegeneration and functional outcome. In contrast, interanimal variability in the present study provided us a valuable spectrum of changes that made it possible to assess the association between the severity of damage in histologic preparations and MRI findings.
It is unlikely that the volume reduction found in this study in the various temporal structures is due to growth retardation. An 8% difference in body weight was found between controls and experimental animals at the time of weaning. Previous studies on the influence of malnutrition on growth of body and brain demonstrated that decrease of body weight by more than 25% did not affect the wet weight of the brain of undernourished rats during the preweaning period (17).
SE at P12 in rats causes neurodegeneration in the medial temporal lobe
MRI analysis of the hippocampus 3 months after SE indicated that the mean thickness of the septal hippocampus did not differ between controls and rats with SE at P12. This was in accord with histologic analysis of the same sections. A more detailed histologic analysis of different hippocampal subfields, however, indicated mild decrease in the thickness of the CA1 subfield in rats with SE. Consistent with the present observations, SE-induced acute hippocampal neurodegeneration in selected subfields has been demonstrated after prolonged seizures in immature (P10) rats in several previous studies (18–20). Neurons were considered as “injured” rather than “degenerating” because authors did not observe neuronal loss in long-term samples. Consistent with previous observations, we did not detect any clear neuronal loss in the CA1 pyramidal cell layer or the dentate granule cell layer in visual scanning of Nissl-stained chronic sections 4 months after SE. This was surprising because most of the animals likely had degenerating neurons in the CA1 at the acute phase. Further analysis, however, indicated that Fluoro-Jade B–positive neurons encompassed ≥50% of the length of the CA1 in only one of five rats. Probably a more substantial acute cell death is needed to detect laminated hippocampal neurodegeneration in routine histology of chronic samples.
SE-induced damage in the CA1 hippocampal subfield was previously described in 2- to 3-week-old rats (14,15). In 1-week-old rabbits, Franck and Schwartzkroin (21) described loss of pyramidal neurons in the CA1 after systemic injection of kainate. They hypothesized that lesion to the CA1 resembles that induced by hypoxia or ischemia or both. More recently, however, in a model of LiCl/pilocarpine-induced SE in P10 rabbits, Thomson and Wasterlain (22) demonstrated vulnerability of CA1 neurons even though oxygen availability remained unchanged during 1-h SE. Therefore they suggested that CA1 damage is likely to be due to seizure activity rather than to hypoxia.
As our data show, acute neuronal damage in the hippocampus of rats with SE at P12 can be detected as reduction in hippocampal volume at the chronic phase by MRI or hippocampal morphometry. Unlike that in animals with SE at an older age, chronic reduction in hippocampal volume in rats with convulsive SE at P12 is not associated with laminated hippocampal cell loss that would be easily detectable in routine histology. Another important aspect of the analysis is that the measurement of hippocampal volume or thickness included not only neurons but also extracellular space, glia, axons, dendrites, and blood vessels. For example, Jiang and collaborators (23) found an almost 30% reduction in the dendritic spine density, and a 40% decrease in the diameter of dendritic arbors in CA3c pyramidal cells after recurrent seizures induced by intrahippocampal injection of tetanus toxin in P10 rats. Thus it remains to be studied whether pathologies other than loss of neuronal somata contribute to volume reduction in immature hippocampus exposed to SE at early stages of postnatal development.
Based on MRI volumetry, amygdaloid damage is present in a subpopulation of patients with TLE with or without SE as an etiology for epilepsy (24). The present data demonstrate that convulsive SE can cause amygdaloid damage at P12 that can be detected at 3 months after SE, both histologically and with MRI. Scattered degenerating neurons were found in both the amygdala and piriform cortex 8 h after SE. Furthermore, a 13% reduction was seen in total neuronal number in layer III of the piriform cortex 2 months after LiCl/pilocarpine-induced SE, as estimated by unbiased stereology (Kubova and Pitkänen, unpublished data). It is likely that neurodegeneration in the piriform cortex and the amygdala contributed to the volume reduction in the amygdalopiriform region in MRI analysis. It remains to be explored whether amygdaloid damage is associated with emotional impairment in subjects with early-life SE.
MRI volumetry reveals atrophy of the perirhinal cortex in ∼30% of patients with drug-refractory TLE (25). In adult rats, SE-induced damage in the perirhinal cortex has not been described in detail, but our unpublished observations indicate neurodegeneration in layers II, III and VI (Lukasiuk and Pitkänen, unpublished data). Here we show that SE at P12 can cause scattered neuronal degeneration in area 35 of the rat perirhinal cortex that can be detected in histologic analysis acutely and in MRI 3 months later. The contribution of perirhinal neurodegeneration to memory performance of subjects with early-life SE remains to be studied.
Hippocampal and amygdaloid volume reduction is present in a subpopulation of adult rats with SE in early life
As in humans, only a subpopulation of rats exhibits volume reduction in the medial temporal lobe after SE. MRI analysis of the temporal lobe 3 months after SE indicated that 23% of the rats had hippocampal, 15% amygdaloid, and 31% perirhinal volume reduction. Histologic analysis of sections from the MR-imaged brain was in good agreement with the MRI data. Analysis of neurodegeneration immediately (8 h after SE) indicated that damaged neurons were observed in the same temporal lobe regions as the volume reduction in chronic samples. Interestingly, in three of five animals, most of the degenerating neurons were in the amygdaloid complex, whereas in two of five animals, a similar density of degenerating neurons was noted in both hippocampal and extrahippocampal areas. These data indicate that only a subpopulation of rats with SE at P12 develops neurodegeneration and temporal lobe volume reduction in long-term follow-up. Further, the distribution of neurodegeneration varies from case to case. These observations are of interest regarding our recent analysis, which indicated that ∼25% of rats with LiCl/pilocarpine-induced SE at P12 develop epilepsy in 3 months, and another 40% have interictal spiking (6). Further studies are needed to clarify whether early abnormalities in MRI can be used as surrogate markers to associate structural damage with epileptogenesis and poor cognitive and behavioral outcome.
SE at P12 leads to neurodegeneration in the medial temporal lobe. Unexpectedly, the distribution of degenerating neurons in the hippocampus was scattered rather than laminated. Dying cells in the amygdala also were not clustered into some subnucleus. The present data strengthen the view that even though SE can induce neurodegeneration as early as P12, factors other than major neuronal cell death might have a significant role in the development of brain-volume reduction and epileptogenesis after early-life SE (26,27). The contribution of mild neurodegeneration to other outcome measures, like developmental delay or cognition, requires further study.
- Top of page
Acknowledgment: We gratefully acknowledge the expert technical help by Mrs. Merja Lukkari (Kuopio, Finland) in histology and by Dr. Jirmanova (Institute of Physiology, Prague, Czech Republic) in electron microscopy. The help of Dr. Riitta Miettinen (Kuopio, Finland) in interpretation of electron microscopic images is greatly appreciated. The present study was financially supported by the Exchange Visitor Program between the Academy of Sciences of the Czech Republic and the Academy of Finland (to H.K. and A.P.), grant 304/05/2582 of the Grant Agency of the Czech Republic (H.K.), Research Project AVOZ 50110509 (H.K.), the Academy of Finland (A.P.), and the Sigrid Juselius Foundation (A.P. and R.A.K).